[05:42] morning [07:08] morning [07:10] pstolowski: hey [07:11] mvo: hey, the last day of sprint! [07:19] mborzecki: good morning! [07:19] mborzecki: yeah [07:29] PR snapd#8979 closed: tests: more checks in core20 early config spread test [08:01] zyga wrote me he's off today [08:24] mborzecki: I think this might be the root cause of that parallel install problem reported on the forum: https://forum.snapcraft.io/t/parallel-instances-cant-remove-cant-connect-interfaces-cant-use-layouts-env-vars-are-wrong/18696/8?u=jamesh [08:25] i.e. whether a non-instance keyed version of the snap is installed or not [08:25] jamesh: we ensure that my-snap directory exists when my-snap or my-snap_* gets installed [08:26] jamesh: or at least we did and had some tests for it, let me double check the tests are actually testing the right thing [08:27] jamesh: fwiw, quick install of a random snap with instance key creates the snap-name (without instance key) directory [08:28] mborzecki: it certainly looks like the $SNAP mount is failing somehow, based on his message [08:40] jamesh: asked more questions there, i suspect it's something else being broken here [08:40] sadly zyga is off today [09:45] PR snapd#8995 opened: osutil: add CheckFreeSpace helper (1/N) === pedronis_ is now known as pedronis [11:45] seriously it's hard to tell when soemone is trolling or being misinformed [11:49] running spread tests for https://bugs.launchpad.net/ubuntu/+source/snapd/+bug/1887134 [11:49] Bug #1887134: Purging of snapd package should remove all files in / installed by snaps

[11:49] quick errand, back in 30 === Eighth_Doctor is now known as Conan_Kudo === Conan_Kudo is now known as Eighth_Doctor [12:19] do we have anywhere in the snapd codebase where we wait for systemd units to become active ? [12:21] mborzecki: do you know ? [12:21] pstolowski: or maybe anything we do in managing snap services? I seem to remember we did something like this somewhere but I can't find it [12:25] ijohnson: one sec [12:25] ijohnson: let me check, iirc we had --no-wait for some actions [12:25] haha thanks folks [12:25] ijohnson: but systemctl start waits for unit to become active (where the meaning of active depends on the type of unit) [12:25] there's systemd.IsActive [12:26] that's a thin wrapper around systemctl is-active [12:27] mmm so actually I'm using systemd-mount from the initramfs [12:27] and looking at the docs it says this [12:27] > the job will be verified, enqueued and systemd-mount will wait until the mount or automount unit's start-up is completed. [12:28] it's not clear to me though that means the unit is active, it kinda seems like that means the unit exists, not necessarily that the mount is done [12:28] ijohnson: there's StartNoBlock but it's used only for snapd services on core18+ [12:29] ijohnson: iirc for mounts we have some special handling in link.go that double checks that snap metadata is present at the mount location [12:29] mmm ok, yeah it seems like we don't have what I thought we had [12:29] thanks for looking! [12:30] ijohnson: err, it's in snapstate/handlers.go in doMountSnap [12:30] ijohnson: there's a for loop that retries readInfo right after SetupSnap [12:46] PR snapd#8996 opened: packaging, cmd/snap-mgmt, tests: remove modules files on purge [13:16] cmatsuoka: what input the new split helper fails on? [13:16] mborzecki: ==a, a==, maybe a lone =? [13:17] haha [13:17] let me try that [13:17] I mean, it's not that it fails, but I don't know what the kernel thinks about that [13:23] cmatsuoka: me neither, probably ignores that because what else can you do at this point [13:24] probably it will ignore all ill-formed parameters [13:24] well, or maybe parse it partially, who knows [13:24] s/it/them/ === plars_ is now known as plars [13:59] PR snapcraft#3209 opened: extensions/desktop/common: add snapd gl vdapu dir to LD_LIBRARY_PATH [14:01] ijohnson, ^^^ that should be vdpau not vdapu ... [14:02] man I open one PR to snapcraft and 2 people tell me it's got a typo within less than 2 minutes [14:03] teamwork ! [18:07] PR snapd#8997 opened: packaging: add "ca-certificates" to build-depends [18:12] PR snapd#8998 opened: tests/cmd/snap-bootstrap/initramfs-mounts: fix typo [18:55] PR snapcraft#3210 opened: docker: install snapd dependency [19:42] PR snapd#8997 closed: packaging: add "ca-certificates" to build-depends [19:57] PR snapd#8983 closed: tests: add tests.cleanup helper [23:10] Bug #1887204 opened: Snap is not clear about partial strict confinement [23:13] Bug #1887204 changed: Snap is not clear about partial strict confinement [23:19] Bug #1887204 opened: Snap is not clear about partial strict confinement